Measurement of tumor blood flow and metabolism by non-invasive methods, such as PET, is key to following the effect of anti-angiogenic treatment in tumors. Preliminary result obtained by PET in our phase I study of anti-angiogenic treatment with Endostatin, have yielded two critical pieces of information. The first is that at low doses of Endostatin (<60 MG/M2), both blood flow and glucose metabolism are increased as a function of time. At higher doses (>120 mg/m2), there is a clear reduction of blood flow to the tumors but an increase in glucose metabolism. These results suggest that blood flow decrease may precede metabolism decrease in anti-angiogenic treatment of tumors and that tumor blood flow, and glucose metabolism become uncoupled under moderately ischemic conditions. The second is that tumors can change in size, shape, and distribution, very rapidly. The imaged lesions are heterogeneous in blood flow and metabolism and therefore an accurate measure of these crucial parameters will require the development of new and automated methods of data analysis that are more sensitive and reproducible. Based on these issues, our Specific Aims in this Developmental Project are: (1) To define the parameters that will provide insight into the sequential changes occurring in blood flow and metabolism in tumors following therapy with anti-angiogenic agents. (2) To develop methods and software tools to automatically extract this information from serial PET images of human tumors with emphasis on reproducibility and accuracy. (3) To validate these methods using data from a phase I trial of Endostatin; with subsequent application to other early clinical trials of putative anti-angiogenic therapies such as PKI166 (Core D) and SU6668 (Developmental Project 4). The short-term significance of this Developmental Project is that these new and automated methods will shorten the time required for early clinical testing of new anti-angiogenic agents. The long-term significance of this project is that these tools may eventually be used in patients to accurately predict anti-angiogenic effects on tumors thereby optimizing anti-cancer effects.